Dell, HPE Post Record AI Server Revenue in Q2

Dell Technologies and Hewlett Packard Enterprise (HPE) have both reported record AI server revenue in their most recent quarterly earnings, underscoring the explosive growth of enterprise AI infrastructure spending. The results mark a turning point for traditional server manufacturers that have successfully repositioned themselves as critical players in the generative AI supply chain.

The combined results suggest that enterprise AI adoption has moved well beyond the hyperscaler-dominated first wave, with corporations across industries now investing heavily in on-premises and hybrid AI compute infrastructure.

Key Takeaways at a Glance

• Dell's Infrastructure Solutions Group (ISG) posted record quarterly revenue, with AI server orders surging past $3 billion in the quarter
• HPE's server segment delivered its strongest quarter ever, driven primarily by AI-optimized systems featuring NVIDIA GPUs
• Both companies reported AI server backlogs that stretch into 2025, indicating sustained demand
• Enterprise customers — not just hyperscalers — are now driving a significant portion of AI server purchases
• Gross margins on AI servers remain under pressure due to the high cost of GPU components
• Dell and HPE are both expanding partnerships with NVIDIA, AMD, and Intel to diversify their AI hardware portfolios

Dell's AI Server Business Crosses a Major Milestone

Dell's earnings call revealed that its AI-optimized server revenue has become a multi-billion-dollar business in record time. The company's PowerEdge server line, particularly models equipped with NVIDIA H100 and H200 GPUs, has seen demand far exceeding initial projections.

Dell CEO Michael Dell highlighted that the company shipped more AI servers in a single quarter than in the entire previous fiscal year. The company's AI server backlog reportedly grew by over 30% sequentially, suggesting demand continues to outpace supply.

What makes Dell's position particularly notable is its strength with enterprise customers. Unlike pure-play AI infrastructure providers, Dell benefits from decades of existing enterprise relationships. Companies already running Dell infrastructure find it natural to extend their investments into AI-capable systems through the same vendor.

Dell also noted growth in its AI-related storage business, as organizations deploying GPU clusters require massive amounts of high-performance storage for training data and model checkpoints. The company's PowerScale and ObjectScale platforms have seen increased attach rates alongside AI server deployments.

HPE Rides the AI Wave With Record Server Revenue

HPE's quarterly results told a similar story. The company's Compute segment delivered record revenue, with AI systems representing a rapidly growing share of total server sales. HPE reported that its AI server orders more than doubled year-over-year.

Antonio Neri, HPE's CEO, emphasized that the company's GreenLake cloud platform is becoming a key differentiator in AI infrastructure deals. GreenLake allows enterprises to consume AI compute on an as-a-service basis, lowering the barrier to entry for organizations that want GPU-powered infrastructure without massive upfront capital expenditure.

HPE's Cray EX supercomputer line has also contributed significantly to AI revenue, particularly in government, research, and large-scale enterprise deployments. The company secured several major contracts for AI supercomputing clusters during the quarter.

Notably, HPE's networking business also benefited from AI buildouts. High-performance networking gear — including switches, adapters, and cables — is essential for connecting GPU nodes in AI training clusters. HPE reported double-digit growth in its networking segment, directly correlated with AI infrastructure deployments.

The Margin Challenge: Revenue Growth vs. Profitability

Despite the headline-grabbing revenue figures, both Dell and HPE acknowledged a persistent challenge: AI server margins remain significantly lower than traditional server margins. The primary reason is the outsized cost of NVIDIA GPUs, which can represent 70-80% of the total bill of materials for an AI server.

This creates an unusual dynamic where revenue grows dramatically, but profitability does not scale proportionally. Both companies are working to address this through several strategies:

• Increasing attach rates of higher-margin components like storage, networking, and management software
• Offering professional services for AI deployment, configuration, and optimization
• Diversifying GPU suppliers by qualifying AMD Instinct and Intel Gaudi accelerators as alternatives to NVIDIA
• Building proprietary software stacks that add value beyond raw hardware
• Pursuing longer-term service contracts that bundle hardware with ongoing support revenue

Wall Street has largely accepted the margin compression narrative, rewarding both stocks for top-line growth and market share gains in AI infrastructure. Analysts argue that winning AI infrastructure relationships today creates long-term revenue streams through upgrades, expansions, and associated services.

Enterprise AI Adoption Enters Its Second Phase

The Dell and HPE results reflect a broader industry shift. The first phase of AI infrastructure spending was dominated by hyperscalers — Microsoft, Google, Amazon, and Meta — building massive GPU clusters for training foundation models. That phase drove NVIDIA's meteoric revenue growth but largely bypassed traditional enterprise server vendors.

The second phase, now clearly underway, involves enterprise customers building their own AI infrastructure. These organizations are deploying AI for specific use cases:

• Financial services firms running proprietary trading models and risk analysis
• Healthcare organizations processing medical imaging and drug discovery workloads
• Manufacturing companies implementing predictive maintenance and quality control
• Telecommunications providers optimizing network operations with AI
• Government agencies deploying classified AI systems that cannot run in public clouds
• Energy companies using AI for exploration data analysis and grid optimization

This enterprise wave plays directly to the strengths of Dell and HPE, which have established sales channels, service organizations, and financing programs tailored to corporate IT buyers. Unlike hyperscaler purchases that often go directly to NVIDIA or ODM manufacturers, enterprise deals typically flow through established vendors like Dell and HPE.

Competitive Landscape Intensifies

Dell and HPE are not the only companies chasing the AI server opportunity. Supermicro has emerged as a formidable competitor, posting its own record revenue figures and gaining market share with a strategy focused on speed-to-market and customization flexibility. Supermicro's liquid cooling capabilities and rapid design cycles have made it a favorite among AI-first customers.

Lenovo is also expanding its AI server portfolio aggressively, particularly in Asian and European markets. The company's ThinkSystem servers with NVIDIA GPUs have gained traction in research institutions and government deployments.

Compared to these challengers, Dell and HPE differentiate through end-to-end solution capabilities. Both can deliver not just servers, but complete AI infrastructure stacks including storage, networking, software, and services. This integrated approach appeals to enterprise customers who lack the in-house expertise to assemble and optimize AI infrastructure from individual components.

What This Means for the AI Industry

The record earnings from Dell and HPE carry several important implications for the broader AI ecosystem.

First, they confirm that AI infrastructure spending is broadening beyond a handful of tech giants. When Fortune 500 companies across diverse industries are purchasing AI servers, it signals genuine enterprise adoption rather than speculative investment.

Second, the results validate the on-premises and hybrid AI deployment model. Despite the dominance of cloud computing narratives, many organizations prefer — or require — local AI infrastructure for reasons of data sovereignty, latency, cost control, or regulatory compliance.

Third, the sustained backlogs suggest that GPU supply constraints continue to shape the market. Even as NVIDIA ramps production of its latest Blackwell architecture, demand appears to consistently outstrip supply, creating long lead times that benefit vendors with established allocation relationships.

Looking Ahead: What Comes Next

Both Dell and HPE provided optimistic forward guidance, projecting continued AI revenue growth through the remainder of the fiscal year and into 2025. Several catalysts could further accelerate demand.

NVIDIA's Blackwell GPU platform is expected to drive a significant upgrade cycle, as customers seek the performance improvements necessary for training ever-larger models. Both Dell and HPE have announced Blackwell-based server platforms and are already taking pre-orders.

The emergence of sovereign AI initiatives — where national governments invest in domestic AI compute capacity — represents another growth vector. Multiple countries have announced plans to build national AI infrastructure, creating opportunities for vendors capable of delivering at scale.

Finally, the rise of AI inference workloads could prove even more significant than training. As enterprises move AI models from development into production, they need inference infrastructure deployed across data centers, edge locations, and even on-premises sites. This distributed deployment model favors vendors like Dell and HPE with broad product portfolios spanning data center and edge computing.

For investors, developers, and IT leaders, the message from this earnings season is clear: enterprise AI infrastructure is no longer a future promise — it is a present reality generating billions in revenue and reshaping the server industry in real time.